It is known to tow, at the end of a cable, behind a boat, a submerged body called a fish because of its resemblance to this animal. Outside these periods of use, the fish is brought back on board the boat by winding up the towing cable on a winch. To facilitate handling, the cable passes over an idling pulley located at one end of an arm articulated at its other end to the boat, in such a way that, by rotating about this articulation, the arm passes from an extended position, in which the pulley hangs out over the sea, to a raised position, in which it hangs out over the deck of the boat. The arm also includes, at this pulley, a device forming a cradle on which the fish comes to rest after it has been raised.
On account of the equipment on board the fish, the latter can reach a considerable mass, of the order of several tones. Since the towing structure moves along only at a relatively low speed, on the order of about ten knots at most, in particular so as not to impose too high a tensile stress on the towing cable, the vertical distance of separation of the fish from the boat is relatively small and the angle between this vertical and the straight line which joins to the boat to the fish is typically on the order of 30.degree..
Moreover, on account of the stresses which it has to withstand, the traction cable is itself large and heavy. Because of this, it is not possible to neglect its weight in relation to that of the fish and it furthermore is subjected to significant drag. What is more, it exhibits significant deflection in relation to the straight line joining the fish to the boat.
Under the effect of the waves, the boat displays random movements which are passed on to the cable and to the fish. Because of the above effects, the movement of the fish does not reproduce that of the boat and it is therefore not possible, during the processing of the sonar signals for example, to compensate for the movements of the fish simply by taking into account the movements of the boat. What is more, the stresses in the cable due to these movements are absolutely excessive and it is necessary to minimize them as far as possible.
A known solution to this problem consists supporting the arm with the aid of a cylinder actuator connected to a pneumatic accumulator, so that the arm oscillates about its point of articulation on the boat so that the head of the arm follows a substantially rectilinear path at an approximately constant speed. As may be easily imagined, this result is obtained only very approximately with such a passive system. In addition, on account of the masses and stiffnesses involved, the natural frequency of this assembly is, in most cases, relatively close to that of the swell, which leads resonance phenomena liable to increase the stresses applied to the cable.